We believe it is likely that therapies for the prevention of type 1 diabetes will be most efficacious if administered before the development of "fixed" expression of anti-islet autoantibodies. We have evidence that we can now identify infants with a high risk of progressing to autoimmunity and diabetes. There are many avenues for primary preventive therapy and the current proposal will illustrate one such trial, which will likely be altered prior to implementation and likely be part of a collaborative effort to test a number of therapeutic options and regimens. Immunologic "vaccination" with insulin or insulin peptides prevents diabetes of NOD mice with the most dramatic protection associated with early therapy. Amongst islet autoantibody positive relatives of patients with type 1 diabetes, pilot trials of subcutaneous insulin therapy suggest that such therapy will delay progression to diabetes. A major determinant of progression to diabetes (despite therapy) appears to severe loss of insulin secretory capacity at trial entry. This suggests that the therapy is being tested late in the disease process. The results of our prospective DAISY study and the BABY-DIAB study from Germany indicate that DR3/4 (DQ8/DQ2) relatives of patients with type 1 diabetes have an extremely high risk of progression to anti-islet autoimmunity approximately (20- 40%) prior to age 2 (and to diabetes in early childhood). We will test the hypothesis that an "autoantigen" based trial for the prevention of anti-islet autoantibodies can be carried out (e.g. ability to identify high-risk relatives, significant progression to autoimmunity, adequate recruitment). The current proposal will initiate a trial of subcutaneous insulin "vaccination" for the prevention of anti-islet autoimmunity in high-risk children. High-risk individuals will be identified by family history and HLA typing. The major outcome variable of the trial will be the effect upon appearance of GAD65 and ICA512/IA-2 autoantibodies. In addition we will prospectively evaluate in the context of the trial specificity and isotype of anti-islet autoantibodies and characterize anti-insulin T lymphocytes, as well as analyze genetic determinants of anti-islet autoimmunity.